Abstract In the classic models of nuclear receptor (NR) action, it was held that the role of ligand was merely that of a switch converting an inactive receptor to one that could positively or negatively regulate gene transcription. However, the discovery of cofactors, coactivators and corepressors, and the observation that the biological responses to hormones and anti-hormones is influenced by the absolute and relative levels of these different proteins, has significantly changed our understanding of NR signaling. It has recently become clear that nuclear receptors (NRs) engage different cofactors in the course of their regulatory activities and that the biological consequences of differential cofactor recruitment are not equivalent. The challenge that has arisen therefore is to (a) define the cofactors used by each receptor in each of the different activities in which it is engaged within the cell and (b) evaluate how in pathologies such as cancer these cofactors are used for the benefit of the tumor. It was within the background of this important issue that we initiated a project, during the previous funding period, aimed at identifying the cofactors that interact with AR and defining how these factors enable the receptor to manifest different activities in different cells. More specifically, we were interested in identifying the proteins utilized by AR when it is engaged in processes required for the growth, survival and metastasis of prostate cancer cells. We have recently completed a comprehensive screen for AR interacting proteins expressed in androgen- responsive tissues that yielded close to 300 cofactors. A chemical biology approach was used to demonstrate that these cofactors can be separated into different classes based on the surfaces on the receptor with which they interact. The primary goal of our continued efforts in this area is to define the specific roles of these different cofactors on androgen-regulated processes elaborated in cellular models of prostate cancer that track with disease progression. To accomplish these goals we propose the following specific aims: (1) Definition of the biological pathways regulated by androgens in cellular models of prostate cancer that are manifest in prostate tumors and track with progression, (2) Identification of the cofactors utilized by AR in processes linked to prostate cancer pathogenesis, and (3) Definition of the roles of specific AR cofactors in the growth, survival and metastasis of prostate cancer xenografts. We anticipate that the results of these studies will be informative with respect to androgen signaling in prostate cancer cells and may shed light on the molecular events that allow cells to transition from a hormone-dependent to a hormone-independent state. Additionally, the information obtained may also assist in the identification of targets with which novel therapeutics can be developed for the treatment of prostate cancer.